373 Emergency Physicians Can Safely Treat and Release Selected Influenza, Pneumonia and Pyelonephritis Patients With Sepsis

Abstract

Some patients (pts) with “sepsis”, as classically defined, are nonetheless “treated and released” (Tx/Rel) from the emergency department (ED). [“Sepsis” means at least 2 “SIRS” criteria with an infectious diagnosis (dx)]. Careful ED physicians (EDPs) probably use “physician gestalt” to identify septic pts for this strategy. Prior peer-reviewed outcomes studies of septic pts have included only those hospitalized. Thus, no peer-reviewed literature exists to support an ED “Tx/Rel” strategy for any septic pts. We tested the hypothesis that the 95% CI for 7-day (d) mortality of septic pts Tx/Rel from the ED with a dx of pneumonia (PNEU), influenza (INF) pyelonephritis (PYELO), Urinary Tract Infection (UTI) or undifferentiated sepsis (SEP) includes 0%. We tracked the short term outcomes (STO) of 7-d & 30-d mortality, & 7-d & 30-d returns for ED care. Search of the Social Security Death Index verified a lack of short-term (<30 d) mortality among pts who did not return to our ED for any subsequent care. SIRS scores, not qSOFA scores, plus one of these 5 infectious dx, were used to identify the study group with sepsis, because qSOFA score is a prognostic tool, validated in an Intensive Care Unit (ICU) setting, & because a valid prognostic study requires all pts to have a correct dx upon study entry. Thus, qSOFA is a tool unsuitable for identifying septic pts in the ED, because: 1) Use of qSOFA for diagnostic purposes would require “circular reasoning” & 2) An ED environment differs from ICUs where qSOFA was validated. *Between 1/1/2017 and 12/31/2018, 2677 patients were Tx/Rel from a busy tertiary hospital ED (2018 pt volume 59,788) with a dx of PNEU, INFL, PYELO, UTI or SEP. Of these, 1067 patients were “SIRS-positive” at ED arrival. *The 7-d and 30-d mortality of these 1067 pts was zero (95% CI 0-0.28%). Thus, these pts were safely Tx/Rel. 4/1067 returned to the ED within 30d for an unrelated reason. 17/1067 returned with a chief complaint (CC) apparently related to the 1st visit, but without evidence of decline of clinical status; 13/17 within 7d. All 17 were again Tx/Rel. 23/1067 pts returned to the ED within 30d due to evidence of decline or failure to improve as expected, of whom 1 returned beyond 7d. All 23 had a change in the therapeutic plan implemented, such as administration of oxygen and/or change of antibiotic. 13/23 were admitted to hospital at time of the 2nd ED visit, none to an ICU, & 12/13 within 7d. 10/23 were again Tx/Rel. Characteristics (Age, Sex, Dx, SIRS & qSOFA scores of those 23 with objective evidence of need to return to the ED will be presented at the Forum. Only 9/1067 SIRS+ pts had initial qSOFA scores of 2 or 3. ED physicians can safely Tx/Rel selected septic patients, apparently via “physician gestalt,” with very low risk of short-term mortality. Occasionally, septic pts Tx/Rel from an ED will deteriorate or fail to improve after discharge, demonstrating a need for clear “ED return precautions.”